On 4 November the US National Association of Wine Retailers called a major alcohol study biased and manipulated, and said it buries the good news about moderate wine drinking. The paper in question, the Alcohol Intake and Health Study, came out in the Journal of Studies on Alcohol and Drugs in June. Its title says there is no protective effect at low levels, with mortality rising to 1 in 25 at 14 drinks a week.
We are a wine site and we sell things to wine lovers, so we have a stake here too. That is exactly why we did not want to pick a side from the press releases. We read the paper, tables included.
This is not medical advice. It is a look at what the numbers do and do not say.
What the study did
It is a model, not a trial. The authors combined published risk estimates for the many diseases and injuries linked to alcohol with US data on how much people drink and what they die of. The result is a lifetime risk of dying from an alcohol-caused condition, compared with people who never drank.
Here is the core of it, with US standard drinks turned into bottles of 13% wine. A US drink is 13.6 grams of alcohol, so a bottle holds about 5.7 of them. These are lifetime averages, not one bad week.
| Drinks a week | Roughly, in bottles | Men, per 1,000 | Women, per 1,000 |
|---|---|---|---|
| 1 | a fifth of a bottle | -16.3 | -14.8 |
| 3 | half a bottle | -10.6 | -10.9 |
| 7 | 1.2 bottles | 1.7 | 0.15 |
| 14 | 2.5 bottles | 39.3 | 40.5 |
| 21 | 3.7 bottles | 68.9 | 82.4 |
A minus sign means the model’s central estimate is fewer alcohol-related deaths than among people who never drank. Keep that in mind, because it is where the whole argument starts.
The retailers’ figures are real
We checked the numbers in the association’s statement against the paper, and the ones we could check are there.
- Negative figures at 1 to 3 drinks a week: yes, table 1.
- About 8 to 10% less risk of dying from an ischaemic stroke, up to 7 drinks a week: yes, table 2.
- About 26 to 30% less risk of diabetes death for women at 7 to 14 drinks a week: yes, also table 2.
- The “36% lower” risk of heart disease among moderate drinkers who do not binge: yes, from an older review the paper cites.
- A very wide margin around the headline threshold for men: yes, 6.5 drinks a week, with a range from under 1 to 13.5.
- Experts picking their top three reviews, with the highest ranked one going into the model: yes, that is how the paper describes it. “Popularity contest” is the association’s phrase, not a finding.
What the retailers leave out
Those figures come with context that the statement does not dwell on.
The negative totals at low intake are not statistically solid. The range at 1 drink a week for men runs from -46.6 to +11.4, and the authors say plainly the effect was not significant. Almost all of that negative figure comes from deaths at 70 and over, and under 40 the paper finds no protective effect at any level, mostly because of injuries. The stroke and diabetes figures the retailers cite do hold up statistically, but the stroke benefit reverses by 14 drinks a week. The diabetes benefit applies to women, and men show none. The 36% figure sits next to another that matters to anyone who saves it all for Saturday: people with the same weekly average who binge lose the benefit altogether.
The big one, for us, is in the headline. The statement talks about moderate wine consumption. The study does not look at wine. It models grams of alcohol, and the authors list beverage type, drinking with food and speed of drinking among the things they did not evaluate. So it cannot bury a wine benefit, and it cannot rule one out either.
What the title leaves out
The study does not get a free pass either. “No protective effect” is a stronger phrase than its own table. At low intake the central estimates point towards protection, and the ranges cover both benefit and harm. A fairer reading is that the study did not find a net benefit it could be confident in. That is different from finding none.
The authors’ 1 in 1,000 and 1 in 100 thresholds rest on central estimates with wide ranges, as they admit. And the “1 in 25” at 14 drinks a week is itself a range, of roughly 1 in 100 to 1 in 14. They also concede that what you assume about alcohol and heart disease has a big influence on the results, and that the figures describe populations, not individuals. In the other direction, they argue the real risks may be higher than their estimates, partly because people under-report how much they drink.
Then there is the politics. The study was commissioned through a US federal committee to inform the dietary guidelines, and was left out when those guidelines came out in January with no numeric limit at all, just advice to consume less. The association says all six members of the review panel were affiliated with anti-alcohol advocacy groups, citing House Oversight records, which we have not verified. The authors declare no competing interests in the paper. A pro-moderation scientific forum says at least two authors of the earlier draft received funding from Movendi, a temperance organisation. It does not name them, and we have not verified that either. Meanwhile a separate National Academies review linked moderate drinking to lower mortality from all causes than never drinking, and to a higher risk of breast cancer in women.
So: a trade body defending its members’ sales, and public health researchers who openly want tighter limits and higher alcohol taxes. Neither is neutral, and we are not either.
What we take from it
Here is what survives whichever side you favour.
At a small glass a night, the central estimate is small, but the range runs well into harm, and this is not a safe zone the study has found. Above about 8.5 US drinks a week, roughly a bottle and a half, the authors’ central estimate passes 1 in 100, and a generous nightly pour may already be there. After that it keeps climbing. Liver disease and several cancers, breast cancer among them, are the part that no amount of argument over heart disease makes go away.
How you drink matters too. Avoiding big sessions is the one piece of advice that both the study and the retailers’ own figures point towards.
Nobody should start drinking for their health, and nobody should pretend that the wine trade has been cleared. Enjoy your glass because you enjoy it, drink a little less than you think you can get away with, and if you have a health question, ask your doctor rather than a wine site or a trade association.
Sources: the study in the Journal of Studies on Alcohol and Drugs, the retailers’ statement via Food & Beverage Magazine, Wine Industry Advisor, ScienceDaily, International Scientific Forum on Alcohol Research, Food Business on the 2025-2030 guidelines
Hero photo: bottle and glass of red wine, by congerdesign, CC0, via Wikimedia Commons.
